ABSTRACT A prototype, knowledge-based, greenhouse climate control and production management system capable of supporting dynamic optimization and continuous monitoring of the greenhouse environment was designed. The prototype was built using object-oriented programming and design techniques. Basic concepts in object-oriented programming are presented as well as the architecture and operation of the prototypic system. The latter is composed of a semantic network representing a typical greenhouse. The various components of this network are presented and their role are explained. The devised architecture allows the coding of generic, knowledge intensive and dynamic control strategies which make use of both rule-based and object-based knowledge representation paradigms. It provides the hooks and mechanisms needed for system growth and is adaptable to a wide variety of physical environments. Several of the conceptual and cognitive structures built into the system are presented and discussed in terms of a set of design and operating criteria. The object-oriented paradigm was found to be well suited to the problem area.